Multi-Regional Large-Scale Brain Systems

The third lecture will be devoted to computational modeling that goes beyond local circuits towards muti-regional large-scale brain system modeling. This is becoming possible only now, thanks to the availability of novel technological advances in anatomy (connectomics) and neurophysiology (optogenetics and neuropixel recording). I will present a dynamical model of large-scale macaque cortex based on recently published weighted- and directed- inter-areal connectivity matrix. A hierarchy of timescales naturally emerges in this model, which is functionally desirable: early sensory areas operate on short timescales appropriate for rapid processing of stimuli, whereas higher association areas display slow dynamics suitable for time integration in decision-making. A key notion emerging from this research is that of macroscopic gradients, namely cellular and connection properties vary systematically along certain preferred axis across the entire cortical mantle. This general principle of large-scale cortical organization, supported by analysis of genetic expression patterns in human as well as mice cortex, will be highlighted.